An oligonucleotide probe for the detection of hepatitis B virus DNA in serum

Journal of Virological Methods, Elsevier

15 (1987) 139-149

139

JVM 00553

An oligonucleotide probe for the detection of hepatitis B virus DNA in serum Hsiang ‘Clinica/ Biochemistry

Ju Lin i, Pui-Chee

Wu2 and

Ching-Lung

Lai3

Unit, 2Department of Pathology, and ’ Department of Medicine, Hong Kong, Hong Kong (Accepted

29 September

University of

1986)

A novel and practical assay for the detection of hepatitis B virus (HBV) DNA in serum is described that utilizes as probe a 21-nucleotide sequence 5’-d (CTTCGCTTCACCTCTGCACGT) labelled at the 3’-end with [s’P]ddAMP. The oligonucleotide probe sequence occurs in all known HBV genomes and is complementary to a region near the end of the single-stranded gap. It includes the 11-nucleotide direct repeat 5’-d(TTCACCTCTGC). The method was tested on 988 serum HBsAg-positive or -neg ative specimens and compared to results with HBV DNA probe, with over 98% concordance between the methods. The sensitivity of the two assays was comparable. The assay was developed for testing serum samples fixed to nylon or nitrocellulose membranes. Hybridization time could be shortened to a few hours as compared to 16 h for HBV DNA probes. Immaculate backgrounds were obtained by using a hybridization medium containing polyethylene glycol, heparin and pyrophosphate, and a particular washing procedure. Serum hepatitis B virus DNA; probe; Radioautography

Oligonucleotide

probe;

3’-End

labelling;

Molecular

hybridization;

DNA

Introduction Hepatitis B virus (HBV) infection is common in many parts of the world, affecting as much as 15% of some populations (Lai et al., 1984). Screening for serum HBV DNA in such areas is important because it is one of the most sensitive tests for the infective form of the virus (Scotto et al., 1983). This report gives a new method for testing serum for HBV DNA using a custom synthesized oligonucleotide, which does away with the need for cloned HBV DNA or for the preparation of HBV DNA from plasma for use as probes. The sequence chosen, 5’d(CTTCGCTTCACCTCTGCACGT), is complementary to the L(-)-strand in the region of the single-stranded gap. It corresponds to the sequence of the S(+)-strand immediately before its fixed 5’-end. It is found in all HBV genomes of different

Correspondence

to; H.J.

Lin, Clinical

Biochemistry

Unit,

University

of Hong

Kong,

Hong

Kong

140

subtypes that have been cloned. The subtypes and position numbers are: adr, 1456-1476 (Fujiyama et al., 1983); adw, 1584-1604 (Ono et al.., 1983); adw2, 2993-3013 (Vatenzueia et al., 1981); and ayw, X01-1581 (Galibert et al., 1979). The oligonucleotide probe is proposed as an alternative to HBV DNA probes in testing serum for HBV DNA because it is as sensitive but simpler to use.

Materials and Methods Preparation of membranes

Nylon and nitrocellulose membranes (10 x 15 cm) were obtained from Amersham International (U.K.) and samples were applied with the aid of a plastic manifold. Samples consisted either of extracts of serum prepared with proteinase K and phenol (Lin et al., 1986a) or of serum prepared as follows: 25 ~1 serum was mixed with 20 ~1 5% Nonidet P-40-1.5% 2-mercaptoethanol-O.O02% bromphenol blue; after 10 min, 135 ~1 of 0.667 M NaCl-0.667 M NaOH was added and mixed and 170 l.~lwas applied to the membrane (Harrison et al., 1985). After filtration, the membrane was soaked in 200 ml 6 X SSC for 20 min and air-dried (1 x SSC is 0.15 M NaCl-0,015 M Na citrate, pH 7.5). Samples were bonded to nylon membranes by ultraviolet irradiation (20 min). Nitrocellulose membranes were vacuum-dried (2 h, 80°C). Membranes were placed in 6 x SSC containing 0.2% each of bovine serum albumin (BSA), Ficoll and polyvinylpyrrolidine (PVP) for 16 h (63”C), blotted and stored in sealed plastic bags. Nylon membranes could be reprobed after removal of the previous probe (Lin et al., 1986a). Probes

The 21-mer was custom synthesized by CellTech (U.K.). 3’-End labelling was carried out with the aid of a kit (Amersham International) using per 100 ~1: 48 pmol 21-mer, 320 l.rCi ddATP (dideoxyadenosine 5’-[a-32P]triphosphate, specific activity about 5000 Ci/mmol) and 10 l.~lenzyme solution (2 h, 37°C). Oligonucleotides and ddATP were separated on a 20-cm column of Sephadex G-25-150. No attempt was made to separate the 22- and 21-mers. The specific activity was about 10’ dpmipg (7 x 10h dpm/pmol). The procedures for the isolation of HBV DNA from HBsAg-positive plasma, and nick translation of unmodi~ed HBV DNA to obtain probes with about 2 x 10y dpm/pg DNA have been given (Lin et al., 1986a). Total human placental DNA and salmon sperm DNA were similarly nick-translated. Recommended

assay

Serum samples were applied to membranes directly, as detailed above. The membrane was placed in a fresh plastic bag (Ziploc, Dow Chemical Co., IN, U.S.A.) with 4 ml of hybridization mix. The mix contained per ml: lo7 dpm 3’-

141

end-labelled probe, 333 l.1.1NETFAP (2.7 M NaCl, 0.018 M EDTA, 0.54 M Tris (pH 7.8) and 0.3% each of Ficoll, BSA and PVP), 300 p-120% (w/v) polyethylene glycol, 100 ~1 heat-denatured salmon DNA (200 kg/ml), 10 ~1 heparin solution (50 mg/ml, dissolved in 0.1 M NaC1-0.0004 M EDTA-0.006 M Tris, pH 7.4), 10 ul 10% Na pyrophosphate and 30 ~1 10% sodium dodecyl sulfate (SDS). The [Na+] was 1.03 M. The heat-sealed bags were sandwiched between two flat plates topped with 800 g of weights. The whole assembly was held at 63°C for 2-16 h. The membranes were washed five times at 50°C with NEPS (1 M NaCi, 0.01 M EDTA, 0.05 M disodium phosphate, 0.5% SDS, pH 7) over the next 20 h, and then shaken at room temperature with 0.2 x SSC-0.1% SDS for 10 min. In each washing step, 100 ml of fluid was used per membrane. The membranes were air dried, placed between cellulose acetate or polypropylene sheets and exposed to X-ray film with intensifying screens for 22-46 h at -70°C. Other materials and methods

Serum specimens were obtained from hospital staff undergoing routine serological screening for hepatitis B, and from patients with a variety of HBV-related diseases. The methods for detecting hepatitis B surface and e antigens (HBsAg and HBeAg) have been given (Lin et al., 1984). The methods of hybridization with nick-translated probes and nuclease Sl digestion have been described (Lin et al., 1986a). Hybridization with HBV, human or salmon DNA was carried out with 5 X lo6 dpmiml. 5’-End labelling of the oligonucleotide was performed using per 100 ul: 100 pmol fy-32P]ATP (specific activity >5000 Cilmmol), 20 pmol 21-mer and 50 units of T4 polynucleotide kinase in Tris-Mg-dithiothreitol-spermidine-EDTA buffer (Maxam and Gilbert, 1980). Separation of the oligonucleotides from ATP was carried out as described for the 3’-end-labelled preparations. In other experiments, hybridization with the oligonucleotide probe was carried out in a medium containing 10% dextran sulfate, 0.5% Nonidet P-40, 0.9 M NaCl, 0.18 M Tris-HCl (pH 8.0)) 0.006 M EDTA and 0.1% each of BSA, Ficoll, PVP and SDS (Conner et al., 1983).

Results Specificity and sensitivity of the probe

The oligonucleotide probe did not react with human DNA, as shown by probing of membranes to which 72 HBsAg-negative sera were applied in singleton, and 8 control specimens were applied in duplicate (Fig. 1). Most serum specimens contained human DNA, as proven by the radioautogram at the top left. Negative controls for HBV DNA were placed at (A) 1, 2, 7, 8 and (H) 3, 4, 11, 12. These dots did not hybridize to HBV DNA nor to the oligonucleotide probe (top right and bottom, respectively). Control serum specimens that were positive for HBV DNA were placed at (A) 5, 6, 11, 12 and at (H) 1, 2, 9, 10. All of these dots gave po-

142

143

Probe: 22-mer

I) . a

* b

a

Probe: salmon DNA

a

b

* b Probe: HBV DNA

a

b

Fig. 2. Effect of washing conditions on binding of the 22-mer. Dot samples were HBsAg-negative senim spiked with (a) HBV DNA or (b) salmon DNA. Top: membranes that were probed with the 22-mer and washed with 6 x SSC at 4°C (left) or with NEPS at 45°C (right). Bottom: hybridization of the dots to homologous DNA probes under standard conditions (see Methods).

sitive results with HBV DNA and with the oligonucleotide probe. Dots (H) 9, 10 reacted strongly with all three probes; they contained large amounts of both human and HBV DNA. The possibility that both the HBV DNA and oligonucleotide probes cross-reacted with human DNA, could be excluded. It was proven that the HBV DNA probes employed did not react at all with samples containing human DNA (Lin et al., 1986a). Moreover, none of the HBsAg-negative serum specimens applied to rows B-G annealed to either the HBV DNA or the oligonucleotide probes. The salient point was that the HBV DNA probe and the oligonucleotide probe behaved similarly to the 80 different serum specimens. The specifity of the 22-mer probe was to large extent determined by the washing conditions (Fig. 2). The samples were HBsAg-negative serum spiked with HBV or salmon DNA. The probe remained bound to both samples after washing with 6 x SSC in the cold. Washing with NEPS at 45°C resulted in binding of the 22mer to HBV DNA but not to salmon DNA. Fig. 3 shows that radioautograms obtained with the oligonucleotide probe or HBV DNA were indistinguishable from each other. The radioactive concentra-

22-mer Probe:

RBV DNA

Fig. 3. Radioautograms obtained with oligonucleotide and HBV DNA probes. Results obtained using the recommended assay with 48 serum samples applied in duplicate to nylon membranes. Hybridization and exposure times were 16 and 22 h, respectively, for both probes.

Probe:

145

tions were 10 x lo6 and 5 x lo6 dpm/ml, respectively. The concentrations were about 10 ng (1.4 pmol) oligonucleotide and 2.5 ng HBV DNA (1.25 fmol) per ml. The time allowed for hybridization (16 h) and exposure time (22 h) were the same for the different probes. Similar comparisons were carried out on serum samples spotted on nitrocellulose membranes with the same results. For ease of handling and convenience, we routinely used nylon membranes and allowed 16 h for hybridization. A comparison of oligonucleotide and DNA probing of 988 serum specimens was made (Table 1). Concordance was obtained in over 98% of the results. The specimens showing discrepant results were borderline cases, being faintly positive with one probe and negative with the other. The reactivities of those 17 specimens were assigned after repeated testing. Of the 174 serum specimens that were positive with both probes, 172 were HBsAg-positive; and of the latter, 127 were positive and 11 were negative for HBeAg. The observed correlation of HBV DNA positivity with the presence of both HBsAg and HBeAg in the large majority of serum specimens was in line with findings in asymptomatic HBsAg carriers and patients with HBV-associated disease (Chen et al., 1986; Gu et al., 1985; Morace et al., 1985). Of the 797 specimens that were HBV DNAnegative with both probes, 50% were HBsAg-negative and 50% positive. In the latter group, 74 were also positive for HBeAg. The absence of HBV DNA in a small proportion of sera positive for both HBsAg and HBeAg has also been recorded (Scotto et al., 1983; Chen et al., 1986; Gu et al., 1985; Morace et al., 1985). Labelling the 21 -mer The 21-mer could be labelled at either end. However, 5’-end labelling was inefficient and expensive since it was necessary to have a 5-fold excess of radioactive ATP over oligonucleotide. 3’-End labelling was satisfactory. It was important to achieve high specific activities because the cold 21-mer competed with the 22-mer probe. Separate experiments showed that hybridization signals could be suppressed completely by adding a 7-fold excess of 21-mer to the hybridization mix.

TABLE 1 Detection of HBV DNA with oligonucleotide

and HBV DNA probes in 988 serum specimens.

Oligonucleotide

probe

Positive

Negative

HBV DNA probe: positive HBsAg HBeAg

174 (172 +, 2 -)” (127 +, 11 -, 36 nt)

9 (9 +, 0 -) (2 +, 4 -, 3 nt)

HBV DNA probe: negative HBsAg HBeAg

8 (7 +, 1 -) (3 +, 3 -, 2 nt)

797 (399 +, 398 -) (74 +, 231 -, 492 nt)

a Number of specimens positive (+), negative (-) or not tested (nt) for the antigen.

146

With the given procedure, 62% (SD 10%) of the 21-mers were labelled and the resulting preparation could be used without separation of the 21- and 22-mers. Time allowed for hybridization Short oligonucleotide probes anneal more rapidly to their targets than nicktranslated probes that are several hundred nucleotides long (Meinkoth and Wahl, 1984). The length of time necessary for hybridization to the oligonucleotide probe studied by comparing the densities of dots hybridized for 0.5, 1, 2, 4, 8 and 16 h. Differences were noted between the 0.5 and l-h samples, while the 1-16-h samples were very similar. Further comparisons showed that a 2-h hybridization period produced dots that were only marginally less dense than those obtained in 16 h (Fig. 4). The hybridization period could, then, be reduced to a few hours. Target site The oligonucleotide probe was in large part bound to the single-stranded region of the viral DNA. Samples of HBsAg-negative serum that were spiked with nu-

Fig. 4. Paired samples probed with the oligonucleotide for 2 and 16 h. Duplicate samples were applied to adjacent lanes, which were cut. Odd-numbered lanes were probed with the 22-mer for 2 h and evennumbered ones for 16 h. The strips were reassembled in their original positions for radioautography.

147

clease Sl-treated or untreated HBV DNA were probed with HBV DNA and reprobed with the 22-mer. Much fainter signals were produced with the oligonucleotide probe in the dots containing nuclease Sl-treated HBV DNA. The fact that nuclease Sl digestion did not altogether abolish binding to the oligonucleotide, suggested a few possibilities: that nuclease digestion might have been incomplete, or that a fraction of the sites complementary to the 21-mer was in double-stranded regions. The probe might also have annealed to the ll-nucleotide direct repeat site in the double-stranded region (Tiollais et al., 1985), but this possibility appeared unlikely, since the 11-base pair hybrid would have a melting point about 30°C lower than that formed with 21 base pairs (Meinkoth and Wahl, 1984). Background

In our hands, a hybridization medium containing dextran sulfate and Nonidet P-40 (Conner et al., 1983) produced patchy and freckled films. The recommended hybridization mixture containing pyrophosphate, heparin (Singh and Jones, 1984) and polyethylene glycol (Renz and Kurz, 1984) was an important factor in producing clean backgrounds. In addition, the final low salt wash was essential. During washing, stacking of membranes was avoided by placing between them spacers cut from plastic flyswatters. Limitations of the assay

The oligonucleotide was not suitable for probing membrane-bound deproteinized extracts of serum that were prepared with proteinase K and phenol, nor for purified nucleic acids that were applied to the membrane without a serum matrix. Nonspecific binding was observed in such samples. It would appear that serum constituents suppressed nonspecific binding more effectively than BSA-Ficoll-PVP treatment, because in radioautograms exposed for long periods of time some HBV DNA-negative dots appeared lighter than the background.

Discussion

The specified oligonucleotide sequence was chosen for several reasons. It was common to all known HBV genomes. Because it corresponded to an S(+)-strand sequence along the single-stranded gap, it would have no competition from viral DNA. With such a probe, it might be possible to detect HBV DNA without having to denature it; however, this possibility was not realized with the dot hybridization technique because the DNA failed to bind to the membranes. It was not foreseen that the oligonucleotide probe would include a sequence of interest in HBV biology. The direct repeat 5’-d(TTCACCTCTGC) is the starting point for transcription of the L(-)-strand (Tiollais et al., 1985). In many instances, integration of the viral DNA occurs in the single-stranded region and portions of the direct repeat appear in the insertions (Dejean et al., 1984; Koch et al., 1984;

Mizusawa et al., 1985; Ziemer et al., 1985). In this connection, it may be possible to use the oligonucleotide probe to search for integrated HBV DNA in biopsy or leukocyte preparations. Such an application of the probe was not studied in this project. Oligonucleotide probes could be universal or selective. The probe we used would appear to be universal. The good agreement between results obtained with the oligonucleotide and HBV DNA probes suggests that sites complementary to the oligonucleotide were present in almost all HBV. However, sequencing, restriction enzyme analysis and molecular hybridization studies have shown that nucleotide sequences of HBV in different carriers may diverge to varying degrees (Charnay et al., 1979; Siddiqui et al., 1979; Lin et al., 1986b). It is possible that other oligonucleotide probes would be useful in characterizing such differences.

Acknowledgements This study was supported by grants from the Pauline Chan Medical Research Fund, the University of Hong Kong, and the Croucher Foundation. References Charnay, P.. Pourcel, C., Louise, A.L., Fritsch, A. and Tiollais, P. (1979) Cloning in E~c~er~c~ia coli and physical structure of hepatitis B virion DNA. Proc. Nat]. Acad. Sci. U.S.A. 76, 2222-2226. Chen, D.S., Lai, M.Y., Lee, S.C., Yang, P.M., Sheu, J.C. and Sung, J.L. (1986) Serum HBsAg, HBeAg, anti-HBe, and hepatitis B viral DNA in asymptomatic carriers in Taiwan. J. Med. Virol. 19, 87-94. Conner, B.J., Reyes, A.A., Morin, C., Itakura, K., Teiplitz, R.L. and Wallace, R.B. (1983) Detection of sickle cell l?&globin allele by hybridization with synthetic oligonucleotides. Proc. Natl. Acad. Sci. U.S.A. 80, 278-282. Dejean, A., Sonigo, P’., Wain-Hobson, S. and Tiollais, P. (1984) Specific hepatitis B virus integration in hepatocellular carcinoma DNA through a viral II-base-pair direct repeat. Proc. Natl. Acad. Sci. U.S.A. 81: 5350-5354. Fujiyama, A., Miyanohara, A., Nozaki, C., Yoneyama; T., Ohtomo, N. and Matsubara, K. (1983) Cloning and structural analysis of hepatitis B virus DNAs, subtype adr. Nucleic Acids Res. 11, 4601-4610. Galibert, F., Mandart, E., Fitoussi, F., Tiollais, P. and Charnay, P. (1979) Nucleotide sequence of the hepatitis B virus genome (subtype uyw) cloned in E. cob. Nature (London) 281, 646-650. Gu, J.R., Chen, Y.C., Jiang, H.Q., Zhang, Y.L., Wu, S.M., Jiang, W.L. and Jian, J. (1985) State of hepatitis B virus DNA in leucocytes of hepatitis B patients. J. Med. Virol. 17, 73-81. Harrison. T.J., Bal, V., Wheeler, E.G., Meacock, T.J., Harrison, J.F. and Zuckerman, A.J. (1985) Hepatitis B virus DNA and e antigen in serum from blood donors in the United Kingdom positive for hepatitis B surface antigen. Br. Med. J. 290, 663-664. Koch, S., Von Loringhoven, A.F., Hofschneider, P.H. and Koshy, R. (1984) Amplification and rearrangement in hepatoma cell DNA associated with integrated hepatitis B virus DNA. EMBO J. 3, 2185-2189. Lai, C.L., Wu, P.C., Yeoh, E.K., Lok, A.S.F., Lin, H.J., Lam, S.K. and Todd, D. (1984) Hepatocellular carcinoma and the hepatitis B virus. In: Viral Hepatitis B Infection (Lam, S.K., Lai, C.L. and Yeoh, E.K., eds.), pp. 3-16. World Scientific Publishing Co., Singapore. Lin, H.J., Wu, P.C., Lai, C.L. and Chak, W. (1984) Micromethod for phosphonoformate inhibition assay of hepatitis B viral DNA polymerase. Clin. Chem. 30, 549-552.

149 Lin, H.J., Lai, C.L. and Wu, P.C. (1986a) Serum hepatitis B viral DNA in HBsAg-positive hepatocellular carcinoma treated with interferon or adriamycin. Br. J. Cancer 54, 67-73. Lin, H.J., Wu, P.C., Lai, CL. and Leong, S. (1986b) Molecular hybridization study of plasma hepatitis B virus DNA in different carriers. J. Infect. Dis., in press. Maxam, A.M. and Gilbert, W. (1980) Sequencing end-labeled DNA with base-specific chemical cleavages. In: Methods in Enzymology, Vol. 65, Part I (Grossman, L. and Moldave, K., eds.), pp. 499-560. Academic Press, New York. Meinkoth, J. and Wahl, G. (1984) Hybridization of nucleic acids immobilized on soiid supports. Anal. Biochem. 128, 267-284. Mizusawa, H., Taira, M., Yaginuma, K., Kobayashi, M., Yoshida, E. and Koike, K. (1985) Inversely repeating integrated hepatitis B virus DNA and cellular flanking sequences in the human hepatoma-derived cell line huSP. Proc. Natl. Acad. Sci. U.S.A. 82, 208-212. Morace, G., Von der Helm, K., Jilg, W. and Deinhardt, F. (1985) Detection of hepatitis B virus DNA in serum by a rapid filtration-hybridization assay. J. Viral. Methods 12, 235-242. Ono, Y., Onda, R., Sasada, R., Igarishi, K., Sugino, Y. and Nishioka, K. (1983) The complete nucleotide sequences of the cloned hepatitis B virus DNA: subtype adr and adw. Nucleic Acids Res. 11) 1747-1757. Renz, M. and Kurz, C. (1984) A calorimetric method for DNA hybridization. Nucleic Acids Res. 12, 3435-3444. Scotto, J., Hadchouel, M., Hery, C., Yvart, J., Tiollais, P. and Brechot, C. (1983) Detection of hepatitis B virus DNA in serum by a simple spot hybridization technique: comparison with results for other viral markers. Hepatology 3, 279-284. Siddiqui, A., Sattler, F. and Robinson, W.S. (1979) Restriction endonuclease cleavage map and location of unique features of the DNA of hepatitis B virus, subtype adw2. Proc. Natl. Acad. Sci. U.S.A. 76, 4664-4668. Singh, L. and Jones, K.W. (1984) The use of heparin as a simple cost-effective means of controlling background in nucleic acid hybridization procedures. Nucleic Acids Res. 12, 5627-5638. Tiollais, P., Pourcel, C. and Dejean, A. (1985) The hepatitis B virus. Nature (London) 317,489-495. Valenzuela, P., Quiroga, M., Zaldivar, J., Gray, P. and Rutter W.J. (1981) The nucleotide sequence of the hepatitis B viral genome and the identification of the major viral genes. In: Animal Virus Genetics (Fields, B., Jalnisch, R. and Fox, CF., eds.), pp. 57-70. Academic Press, New York. Ziemer, M., Garcia, P., Shaul, Y. and Rutter W.J. (1985) Sequence of hepatitis B virus DNA incorporated into the genome of a human hepatoma cell line. J. Viral. 53, 885-892.